So wrote David Hume, in what is now regarded as the locus classicus of the doctrine of monetary “neutrality”.1 According to that doctrine, a change — whether an addition or a subtraction — to a nation’s nominal money stock should, in the long run at least, leave that nation neither better nor worse off than before. That’s so because the change, instead of having any permanent effect on either the extent of employment or the quantities of goods produced by those employed, merely results in an all‐around and strictly proportional (hence “neutral”) change in prices paid for labor and goods. Though it’s perhaps tempting to treat Hume’s insight, and the more modern theory it anticipates, as implying that monetary policies and institutions have little if any bearing upon an economy’s economic growth rate, the temptation should be resisted, because the claim that money is “neutral” actually implies nothing of the sort.
The neutrality doctrine refers, first of all, to the eventual consequences of a one‐time increase in the quantity of money — that is, a multiplication by factor y, or an x‐percent increase. It doesn’t follow that changes in the growth rate of the money stock (or, equivalently, the equilibrium rate of inflation) are also matters “of great indifference.” While economists disagree concerning which inflation rate is best, practically all are in agreement in regarding arbitrary fluctuations in the quantity of money as costly. Such fluctuations contribute to business cycles, whose upward phases involve misallocation of capital as well as some squandering of leisure, and whose downward ones involve wasteful unemployment. Because resources are wasted during boom and bust alike, less stable economies end up having lower growth trajectories than more stable ones.
Finally, the quality of money — that is, the specific assets of which the money stock consists — can have a very considerable bearing upon economic growth, depending on how savings represented by the public’s money balances are invested. In The Wealth of Nations Adam Smith explained how Scotland’s economic growth was aided by the employment, in place of silver coins, of commercial banknotes backed mainly by productive loans. Even today comprehensive currency privatization might raise some countries’ long‐term growth rates by several percentage points, by reducing the extent to which scarce savings are commandeered, and inefficiently invested, by central banks.2
Now let’s consider, in light of these considerations, recent U.S. experience. Before the 2008 crisis the Fed’s liabilities amounted to about one‐eighth of the U.S. (M2) money stock, implying a correspondingly large diversion of savings from potentially more productive investment. Since those liabilities consisted mainly of Federal Reserve notes, many of which circulate overseas, the diversion of U.S. monetary savings was rather less substantial.
Three post‐crisis rounds of Quantitative Easing have, however, led to a fivefold increase in the size of the Fed’s balance sheet. Because banks chose to add the new reserves this easing generated to their exess reserve holdings instead of lending them, the Fed’s share of M2 intermediation rose dramatically, from one‐eighth to one‐third. What’s more, because the Fed’s QE‐related purchases included almost $1.9 trillion in mortgage‐backed securities, such securities now make up well over a third of the Fed’s total assets. Therefore, instead of being invested as productively as possble, or even being used to finance the general government, over one‐sixth of the public’s monetary savings now serve to prop up the value of assets considered toxic by the private sector. Regarded strictly from the perspective of its effect on credit allocation, Quantitative Easing seems to have been very wasteful indeed.
Although the Fed now stands more squarely in the way of healthy U.S. economic growth than ever before, its pre‐2008 performance was, from a growth perspective, also far from ideal. In fact, the monetary environment since the Fed’s establishment has in important respects been even less conducive to growth than that of the previous, crisis‐prone national currency era.3
The price level, for one thing, became far harder to predict, adding considerably to long‐term investment risk. (A stark symptom of this change has been the almost complete disappearance of once‐common 50- and even 100‐year corporate bonds.) Real output has also been more volatile or, if one prefers to overlook the Fed’s first three decades, no less volatile than it was before 1914 — and this despite tremendous broadening of markets, a much‐reduced role for weather‐ and blight‐sensitive farm output, and a much enlarged ratio of government spending to total GDP, not to mention the rise of activist fiscal policy, all of which should have contributed to greater stability. Still more revealingly, shocks to aggregate demand, which would be altogether absent in an ideal monetary system, have become a far greater source of economic instability since 1914 than they were before. Economic contractions, finally, have become more instead of less frequent, as well as longer‐lasting on average, than they used to be — and here once again the comparison holds even setting the interwar period aside.4
From what has been said, it ought to be evident that reducing the U.S. monetary system’s drag on growth means, among other things, (1) minimizing the extent to which that system diverts savings toward relatively unproductive uses, and (2) seeing to it that the system dampens the business cycle instead of aggravating it.
These objectives are, fortunately, not necessarily at loggerheads. For although it is true, as Adam Smith observed, that a fractional‐reserve monetary system is bound to be less secure than one in which all money either is or is backed by basic money, experience shows that the most free and efficient fractional‐reserve systems, including Scotland’s in its heyday (roughly from Adam Smith’s time to the latter part of the 19th century) and Canada’s (to this day, but especially before 1935), have also been among the most stable. Theory suggests, furthermore, that this is neither a coincidence nor a paradox. It is, on the contrary, just what one would expect to find if government regulations tended, as they often do, to make banking systems less rather than more stable than they would be otherwise.5
Reducing wasteful misdirection of savings means, first of all, unwinding the Fed’s bloated balance sheet. Because disposing of its long‐term Treasury and MBS holdings could disrupt the markets for those securities, while exposing it to large capital losses, the Fed is unprepared to tighten money by means of conventional open‐market sales. But should banks become inclined to shed their excess reserves more aggressively, to avoid unwanted inflation, the Fed will have to counter the tendency somehow, which it is most likely to do by increasing interest payments on excess reserves. That strategy would, however, perpetuate instead of ending the present unproductive employment of savings.
This outcome can be expeditiously avoided by two steps. First, the Treasury should be instructed to swap short‐term bills and notes for the long‐term Treasury securities that the Fed is now holding. Second, Congress should establish and fund a special Resolution Authority for the purpose of acquiring the Fed’s MBS holdings at an appropriate discount, and disposing of them in an orderly but relatively rapid manner.
The above reforms would serve to re‐establish something akin to the pre‐crisis status quo. But they would not otherwise diminish the Fed’s share, and hence the overall inefficiency, of U.S. financial intermediation. Further steps in that direction might include doing away with statutory reserve requirements, and (a far more controversial option) allowing commercial banks to issue their own circulating notes, as banks in Hong Kong, Scotland, and Ireland do, and as U.S. banks also did before the Fed acquired its monopoly.
Letting banks issue their own notes makes no sense at all, of course, unless something is first done to correct the moral hazard problems that presently make some banks poor custodians of ordinary deposit liabilities. Such steps must include both a reduction of explicit deposit guarantees and the undoing of the guarantee implied by the official view that some banks are Too Big (or “Systematically Important”) To Fail. Moral hazard stemming from explicit insurance might be contained by providing for co‐insurance, by restricting the uses of (or interest return on) insured deposits (while leaving uninsured balances unregulated), or by replacing government‐administered insurance with private‐industry cross guarantees.6 That stemming from implict insurance can be contained by limiting the Fed’s bailout authority, by having “Big” banks prepare “living wills,” or, if all else fails, by breaking them up.7
Once the moral hazard problem has been addressed, banks may safely be left to intermediate as efficiently as possible, and thus to contribute as much as possible to economic growth, using whatever liabilities their customers might be persuaded to hold, provided that they are protected from such extreme movements in interest and inflation rates as might cause even responsibly managed and well diversified banks to fail. Because such violent swings are almost always due to irresponsible monetary policy, in practice this means placing limits on central banks’ discretionary powers of monetary control.
Limited space prevents me from either rehearsing the arguments or reviewing the evidence favoring monetary rules over discretion, or from considering the many forms such rules might take, from simple base growth‐rate rules to sophisticated ones, such as the Taylor Rule, allowing feedback from various policy targets. I must instead settle for adding my voice to those of Scott Sumner, David Beckworth, and other “market monetarists,” by opining that an ideal rule should seek, not to stabilize either output or inflation per se, or some weighted average of both, but to stabilize some measure of total spending such as the growth rate of nominal GDP or domestic final demand.8 Doing that serves to prevent monetary disturbances from influencing either the general price level or real output, while still leaving both variables free to respond, as efficient allocation requires that they do, to both favorable and adverse supply innovations.
But no monetary rule can succeed that isn’t strictly enforced, and the world has offered no examples — or no long‐lasting ones, at any rate — of a strictly enforced rule since the passing of the classical gold standard. That standard owed much of its success to the fact that central banks actually had relatively little to do either with its enforcement or with its establishment; and there are good reasons for suspecting that the only way to have a truly rule‐bound monetary regime of any sort is by having one that does away with central bankers altogether, instead of merely endeavoring to constrain them. In the Fed’s case, that means replacing the FOMC with some other, automatic means for controlling the money stock.
Because the success of the classical gold standard itself depended crucially on both its international character and the fact that participating banks and central banks still felt themselves bound by contract (rather than by mere “policy”) to maintain their notes’ convertibility into gold, replicating it poses daunting challenges that make it wise to contemplate other options, and especially ones that don’t depend on the cooperation of foreign governments.9
One such option, suggested by Milton Friedman back in the mid 1980s, would be to forego altogether changes to the monetary base, freezing it for all time, and leaving it to private sector financial institutions to accommodate changes in the public’s specific demand for paper currency as well as changes in money’s velocity.10 As I’ve shown elsewhere, if banks are free both to set their own reserve ratios and to issue their own notes, they might accommodate the former changes fully, and the latter to a considerable extent, despite a fixed reserve base.11 Changes in real output would, however, tend to be deflationary; and though the deflation might be harmless or even beneficial to the extent that it reflected productivity gains (it might, for one thing, approximate Friedman’s “optimum quantity of money”), if prompted instead by an increased supply of labor it could instead prove depressing.12
A less Draconian option, and one Friedman also suggested, if only facetiously, would be to replace the FOMC with a computer, programmed to allow the monetary base to grow at a fixed rate. But even that solution couldn’t deal adequately with extraordinary growth in money demand, such as growth stemming from an influx of new workers. It also raises the question: who is responsible for programming the computer, and how can that person or those persons be prevented from tampering with it?
Computer technology has, fortunately, taken huge strides since Friedman made his somewhat tongue‐in‐cheek suggestion; and computer programming has taken even bigger ones. One development especially, and a very recent one at that, seems to me to hold out the best prospect yet for rendering central bankers obsolete. That development, you’ve probably guessed, is Bitcoin.
Bitcoin itself, I hasten to say, has a long way to go before it can qualify as a serious rival to, let alone a replacement for, the U.S. dollar. But it isn’t Bitcoin itself that I regard as a means for dispensing with the Fed and other central banks. It’s the program that controls the supply of Bitcoins — and controls it in such a way as to practically rule out any tampering with that supply.
The actual Bitcoin program or “protocol” provides for a steady but gradually diminishing output of Bitcoins that will level off as their total quantity approaches 21 million (we are, as of this writing, still shy of the half‐way mark). Since this means that the Bitcoin stock will itself one day be “frozen,” this particular protocol, were it used to regulate the supply of base dollars, wouldn’t accomplish anything that simply disbanding the FOMC or shutting down the New York Fed’s open‐market desk wouldn’t accomplish.
But one could also design a “Bitdollar” protocol that, while resembling the actual Bitcoin protocol in being tamper‐proof, allows, not merely for perpetual growth of the bitdollar base, but for growth that automatically responds to changes in, say, the volume of bitdollar payments. One could, in other words, have a “smart” yet tamper‐proof base‐money management algorithm — smart enough, for example, to automatically implement an NGDP rule, or something close to it. Such an automatic system, if only politicians would implement it, offers the best hope yet for monetary stability and, hence, for having a monetary arrangement that contributes to economic growth instead of hampering it.
1 David Hume, “Of Money,” in idem., Essays, Moral, Political, and Literary (Indianapolis: Liberty Fund, 1978).
2 See William D. Lastrapes and George Selgin, “Banknotes and Economic Growth,” Scottish Journal of Political Economy 59 (4) (September 2012), pp. 390–418.
3 George Selgin, William D. Lastrapes, and Lawrence H. White, “Has the Fed Been a Failure?” Journal of Macroeconomics 34 (3) (September 2012), pp. 569–596.
4 See Christina D. Romer, “Changes in Business Cycles: Evidence and Explanations,” Journal of Economic Perspectives 13 (2) (Spring 1999), pp. 13–44. The last conclusion follows provided one allows for events since Romer’s 1999 assessment.
5 See George Selgin, “Legal Restrictions, Financial Weakening, and the Lender of Last Resort,” The Cato Journal 9 (2) (Fall, 1989), pp. 429–59, and Charles W. Calomiris and Stephen H. Haber, Fragile by Design: The Political Origins of Banking Crises & Scarce Credit (Princeton: Princeton University Press, 2014).
6 See Bert Ely, “Financial Innovation and Deposit Insurance: The 100 Percent Cross‐Guarantee Concept,” The Cato Journal 13 (3) (Winter, 1994), pp. 413–36.
7 This solution, though hardly a first‐best ideal, is presumably less far removed from first‐best than it might be when one considers the part government guarantees have played in promoting recent banking‐industry consolidation. Very large banks that have benefited from bailouts (but not those that were compelled to accept TARP money) should top the list of candidates.
8 See Scott Sumner, “The Case for Nominal GDP Targeting” (Arlington, VA: The Mercatus Center, 2012); and William A. Niskanen, “Political Guidance on Monetary Policy,” The Cato Journal 12(1) (Spring/Summer 1992), pp. 281–286.
9 See Selgin, “Law, Legislation, and the Gold Standard,” unpublished, 2014.
10 See Milton Friedman, “Monetary Policy Structures,” The Cato Journal 34 (3) (Fall 2014), pp. 631–55.
11 George Selgin, The Theory of Free Banking: Money Supply under Competitive Note Issue (Totowa, NJ: Rowman & Littlefield, 1988).
12 Idem., Less Than Zero: The Case for a Falling Price Level in a Growing Economy (London: Institute of Economic Affairs, 1997).
The opinions expressed here are solely those of the author and do not necessarily reflect the views of the Cato Institute. This essay was prepared as part of a special Cato online forum on reviving economic growth.